Share of voice is a way of saying how much of the visibility in a set of search results your site has, compared with everyone else who shows up there. This guide is about the organic kind, meaning unpaid Google-style search results, and about the content you publish to earn them. It is a different thing from AI search share of voice, which counts how often a brand appears in the answers that tools like ChatGPT and Perplexity give to a set of prompts. DeepSmith has separate guides on how to calculate AI share of voice and on the tools that measure it, so this one stays with conventional keyword rankings and the pages behind them.
If you have heard the term in a meeting and nodded along, you are in good company. The words are used loosely, and two people can say "our share of voice is 12%" and mean very different numbers. The goal here is to make sure you know what the number is made of, so you can ask the right questions when someone hands you one.
What is SEO share of voice?
SEO share of voice is the portion of visibility, or estimated clicks, that your site earns across a chosen set of organic search queries, compared with the total opportunity or the competitors measured for those same queries.
The part that trips people up is "chosen set." A score built from ten product-category keywords is not your share of every search in your industry. It only describes those ten keywords, in the search engine, country, device and time period you picked. Change any of those and the percentage changes too, so it helps to write them down whenever you report a number.
The term started in advertising, where it described a brand's portion of the ad presence in a market. When we use it for search, we apply the same relative-share idea only to unpaid results. That means it is not the same as paid impression share, social mentions, revenue share or market share, even though people sometimes mix these up in conversation.
A few neighboring terms are worth keeping separate:
- Search visibility score: a weighted summary of your rankings that a tool produces. SISTRIX, for example, describes its Visibility Index as a weighted figure based on organic Google results. A raw index value is not automatically a percentage share, because to turn it into a share you still need a defined set of comparable values and a denominator.
- Share of search: usually the portion of searches for your brand name among searches for competing brand names. It tells you about relative branded demand, which is a different question from how visible you are when someone searches for a topic without naming a brand.
- Rankings, clicks and visibility: a ranking is your position for one query, clicks are the visits a result gets or is estimated to get, and share of voice combines observations across a whole set of queries to express relative presence.
How to calculate share of voice
The general form is simple. Take your measured organic visibility, divide it by the relevant total visibility, and multiply by 100. Almost all of the work is in deciding what goes into each half of that fraction.
Here is the most common way to do it, using search volume as the weight:
- Pick the keywords you want to track.
- For each keyword, estimate the clicks your ranking could get. You do this from the keyword's monthly search volume, your ranking position and an expected click-through rate for that position.
- Add those estimates up across all your keywords. That is your estimated organic clicks.
- Divide by the estimated organic clicks available across the same set of keywords, then multiply by 100.
A simple example makes it concrete. Say the keywords you track offer an estimated 100,000 organic clicks a month, and your site is estimated to capture 6,000 of them. Your share is 6%. Those figures are made up for illustration, and they are not an industry benchmark.
If you want a rougher version to teach the idea, you can use this approximation:
| Step | What you do |
|---|---|
| Estimated clicks for your site | Add up (monthly searches x estimated click-through rate at your position) for each keyword |
| Total opportunity | Add up the monthly searches for the same keywords |
| Approximate share | Your estimated clicks divided by the total searches, times 100 |
Treat that table as a teaching aid and not an exact copy of any platform's metric. Its denominator is searches, while a tool may use estimated available organic clicks instead, since not every search ends in a click on an organic result.
Why weighting matters
One published grocery-search example shows the effect well. It takes five keywords and first averages the per-keyword visibility percentages without any weighting, which gives 7.86%. When it weights the keywords by monthly search volume instead, the result is 8.7%, based on estimated monthly traffic of 410.8 out of 4,720 total monthly searches. The two answers differ because the keywords have different amounts of demand. The example also uses click-through assumptions from its own time, so take it as an illustration of method and not as a current benchmark.
The same thinking explains why simply counting how many keywords you rank for can mislead you. A term with 10 monthly searches and a term with 100,000 count the same in a keyword tally, even though they are not close to being the same opportunity.
The denominator question
There are two common denominators, and they answer different questions. If you divide your estimated clicks by the clicks available across all the tracked searches, you get an opportunity share. If you divide your visibility by the combined visibility of you and three chosen competitors, you get a selected-competitor share.
The second one can go up just because someone removed a competitor from the list, and it can leave out publishers or marketplaces that take up real space in the results. It also does not have to add up to 100% across sites, unless the method spells out how every bit of the opportunity is divided. When you look at any chart, check which denominator it uses before you read anything into the trend.
Using a number a tool gives you
Most teams get their figure from a tool, and that is fine, so it helps to open the tool's own definition of the metric. Vendors differ in their keyword databases, estimated volumes, click-through models, treatment of search-result features, position limits, locations and competitor sets. Moz has a walkthrough on how to calculate share of voice for organic search. Semrush publishes its own walkthrough of how to measure SEO share of voice, and reading two of them side by side shows how much the methods can differ. A percentage from one tool, a percentage from another and an index score from a third should not sit on one trendline as though they were the same thing.
What does content share of voice tell you?
Content share of voice is the same organic visibility question, asked about the topics, keyword clusters or pages your content team is responsible for. It is a useful way of narrowing the scope, and it is not a second, separate formula that everyone agrees on.
The reason to do this is that a sitewide number can hide what is going on. Your product-category keywords and your beginner-education keywords are usually won by different pages, against different competitors, and they can move in opposite directions. If you measure them separately, you can see that your share of the educational topics is dropping while the overall number looks fine, and that tells you where to look.
A cluster here is just an editorial grouping you chose for the analysis. It does not mean your article owns every query in it.
It also helps to keep one thing apart. Counting how many articles you have published, or what fraction of a group's articles are yours, measures coverage or output. That is useful for spotting topics where you have nothing at all, and a topic-cluster plan or a content inventory audit is a good place to start on that. There is also a way to find content gaps without building the analysis by hand. It does not tell you how visible those articles are to searchers, and publishing more pages does not mechanically raise a weighted share. Relevance, whether the pages rank for the measured terms, the demand for those terms and the strength of the pages you compete against all play a part.
If you do want to look at coverage, include topics you have not covered yet in your query set, so the measure does not flatter the site you already have.
What can distort the number?
Most of the trouble with share of voice comes from things that quietly change what the number describes. These are the ones to watch.
A different keyword set. If the keywords, geography, device or competitor list change between two reports, the movement between them is a change in method and not necessarily in performance. Call it a methodology break when you report it.
Branded queries. For a category-level view, leave out both your own brand terms and your rivals' brand terms. Otherwise people who were already looking for a specific company can dominate the result. If branded visibility matters to you, report it as its own number.
Estimates standing in for facts. Rankings, search volumes and expected click-through rates are proxies. They cannot tell you another site's real clicks, so a competitor's estimated share is an estimate and should be described that way. Even your own opportunity denominator is an estimate and not a count of what searchers actually saw or clicked.
The layout of the results page. The same organic position can carry a different click opportunity when the page has extra features around it. A position-only example is a teaching aid and not a promise about click share.
Search Console misread as a competitive measure. Google Search Console is very useful for your own site. Its Performance report shows your property's clicks, impressions, click-through rate and average position, and you can break the table down by query, page, country, device, search appearance and date. It does not show competitors' private impressions and clicks, so on its own it cannot give you a competitive share of voice. Its impressions are also counted in relation to your links appearing in results, and an impression can be counted when a result is on the page without being scrolled into view. Click-through rate is clicks divided by impressions, and average position is an average, not a fixed rank for every search. So neither your impressions nor your average position equals share of voice.
Search presence versus market share. A publisher can win a lot of visibility for a query without selling anything in that category, and a business can sell well without owning the informational results page. It is tempting to carry the old advertising ideas about share of voice and market share over to a single keyword cluster, but they do not work as a prediction there.
No universal "good" number. There is no percentage that counts as healthy across industries. The keyword set, the market and the competition decide what a figure means, so a documented baseline and a comparable trend are more useful than a target someone made up. In the same way, be careful with any single click-through curve by rank that claims to fit every industry, since a current, clearly scoped source is needed before you rely on one.
How should a content team use it?
Use share of voice to decide where to look next, and keep it away from promises about results. Here is a way to set it up that a small team can manage.
- Define the market first. Start with a representative set of queries around the products, problems and topic clusters you can serve. Write down which search engine, location, device and time period you use. Use the same ones for your competitors.
- Decide who the competitors are. Search competitors can include sites that rank for your queries but do not sell what you sell. Decide whether your question is who occupies the results or which direct businesses win visibility, and say which one you picked.
- Look at clusters, not only the sitewide average. Check the share for groups such as beginner questions and product-evaluation queries. Look at which high-demand terms competitors win and which of their pages show up. A sitewide rise can hide a falling share in a cluster you care about.
- Read the number as a signal. A weak share on relevant queries points you toward missing coverage or pages that are underperforming. A rising share suggests your measured search visibility is improving within that fixed set. It does not on its own prove more sales, more brand awareness or more inclusion in AI answers, so look at traffic and business outcomes separately.
- Cross-check with Search Console. Use it to see how your own pages are doing for the queries in your set, and use a rank-tracking or similar competitive dataset for the competitor side of the picture.
- When the share moves, ask what changed. First check whether rankings or estimated clicks changed for the same keywords with the same settings. If the keyword set, location, competitor list or tool model changed, report a methodology break and do not present the movement as pure performance.
Once you have a baseline, a monthly or quarterly check on the same set is usually enough to show whether your content work is moving anything.
Where AI search share of voice fits
Once your organic numbers are steady, you may want to look at how your brand shows up in AI answers too. That is a separate measurement. It watches how often a brand appears in generated answers to a defined set of prompts, and it has its own sampling and its own denominator. A good organic ranking does not by itself establish a particular AI-answer share of voice, and pages that rank well can still miss AI citations, which is why the two are best reported side by side and not blended into one figure. DeepSmith measures AI-answer mention, citation and competitive visibility on its own terms, apart from the keyword-ranking calculation covered in this guide. If you want to go further on that side, our guide to AI visibility metrics is a good next read. A roundup of AI share of voice tools covers the software side. Your old SEO keyword list also will not carry over as a prompt list, which we explain separately.
The short version
Share of voice in search is your slice of visibility across a defined set of queries, and the useful part is that the definition is in your hands. Pick the queries, hold the conditions steady, label your denominator, and look at clusters as well as the sitewide number. Then you have a figure you can compare month to month and explain to someone else in a minute or two.
If you are also working out how your content shows up in AI answers, you can start a free DeepSmith trial and see how it tracks that alongside your content plan.
